CD8+ T cells (TCD8+) play critical roles in limiting viral spread through the elimination of virus-infected cells, which display at their surfaces virus derived peptides in combination with major histocompatibility complex class I (MHCI) molecules. Since many viruses replicate within hours after infection, effective TCD8+ action depends upon catabolism of viral proteins shortly after their synthesis through mechanisms that remain poorly understood. Our pursuit of these mechanisms has revealed that onset of peptide display at the cell surface and, consequently, the density of peptide/MHCI complexes varies considerably depending upon properties of the antigen. Thus, cytosolic proteins and signal sequence (SS)-bearing proteins that are not efficiently translocated into the endoplasmic reticulum appear rapidly while the processing of efficiently translocated proteins is substantially delayed. Furthermore, unshielded hydrophobicity (such as an untethered transmembrane domain) constitutes one driver of sufficiently rapid protein catabolism. Based upon these observations, we hypothesize that TCD8+ specific for epitopes derived from the most rapidly degraded proteins are most protective due to interruption of viral replication at early time points followin infection. Paradoxically, we predict that these same proteins are inferior in priming for a TCD8+ response, since the process of cross-presentation - in which antigen is transferred from an infected cell to the dendritic cell - depends upon stability of the protein. By extension, we propose a strategy for development of a TCD8+-targeting vaccine that involves, first, the identification of a viral protein which, in its natural state, is associated with rapid peptide generation and, therefore more efficient viral clearance. In second step, the antigen is engineered to be stable when expressed by a vaccine, in order to optimize cross- presentation. These ideas will be tested with a combination of in vitro and in vivo experiments using a panel of model antigens expressed primarily by recombinant ectromelia virus, a natural mouse pathogen, and recombinant adenovirus when heterologous expression will be necessary. Results of these experiments could fundamentally alter the rational design of TCD8+-targeting vaccines, a goal with relevance to many human pathogens, including HIV, herpesviruses, and influenza.